Put Content On Every Page
Design shouldn't be mere decoration; it must convey information. A reader should never have to plow through forests of buttons to get simple news.
The fact is that nobody reads anything. The only person that will read every word of what you've written is your mother. Everybody else is too busy. People skim and surf. If you don't give them something quickly, they absorb nothing.
The First Color is White,
The Second Color is Black,
The Third Color is Red
Calligraphers and early printers grasped this over 500 years ago, and experience has proved them exactly right.
On the web, white is every color firing at full strength; it's the brightest color. Black holds the highest contrast to white, and so it is the first choice for type set on a white background. Red (not blue or yellow) works well with both.
Never Letterspace Lowercase
When you do this, the natural, built-in rhythm of the letters is ruined. Despite the current trend in book jackets, this is simply not done. Goudy put it best: "A man who would letterspace lowercase would shag a sheep".
Never Set a Lot Of Text Type in All Caps (Exception: Hunter S. Thompson.)
Fonts were not intended to be set in all caps. They were intended to be upper- and lowercase and have serifs and descenders and ascenders so that they're easier to read. The bottom line is, skip it if it's too hard to read.
A Cover Should Be a Poster
A single image of a human being, preferably Madonna, will sell more magazines than multiple images or all type.
Use Only One or Two Typefaces
Because of the accessibility of fonts, we see designers employing scores of fonts on every screen. But good designs are pulled together with just one or two. The best combination: one light and one bold.
Make Everything as Big as Possible
Type looks great in big point sizes. And a bad picture always looks better bigger.
Get Lumpy
Most design contains no surprise. Page after page of HTML type may be okay if you're running a scientific research site, but if you want normal people to pay attention, change pace. Why do 95 percent of Web sites have a graphic home page followed by legions of pages that look like mimeograph newsletters with snapshots stapled to them? Let's see some ups and downs, some lumps!
Don't Repurpose
Don't dump valid media! Add value, customize for the Web. When we designed USA Today's site, we pushed their flat graphics into wire frames that could be rotated, mapped with new textures, and animated. The content was still there, but it was customized for the Web.
Don't Confuse the Viewer
Your site needs to be consistently designed. If you have different pages and different sections, the navigational tools and graphics need to look the same. Sadly, this is often ignored; sometimes you can't tell if you're still in the same site.
Don't Confuse the Viewer, Part 2
Make sure your buttons and navigational directions are simple and clear. If buttons represent departments, don't add nondepartmental things to the button bar. It's very easy to get confused within a site; make sure yours is painfully clear.
Don't Make Oversize Pages
Over 50 percent of computers have small, thirteen-inch monitors. Everything shrinks significantly from the Web designer's gaudy graphics monitors. Stop kidding yourself: design for 640-by-480-pixel monitors!
Don't Design Pages That Require Scrolling
Just as 75 percent of people will only read the top half of a folded newspaper, most browsers will never scroll. People are much more likely to click a button. Plus, shorter pages also break up the content in easy bite-size pieces.
Don't Use big, Slow Graphics
Nobody wants to wait around for art or video no matter how cool it is. If they have to wait, they'll leave and never come back.
Don't Use a Lot of Colors
Monochromatic pages run faster and look better. If you get your design concept graphically right in black and white, you're on the right track. Add colors cautiously.
Don't Use Blurry Drop Shadows
If we never see another site with blurry drop shadows on every button and every speck of display type, it will be too soon.
Don't Have a Lot of Text
Nobody reads anything anymore. (See Rule 1 of "What to Do on the Web.")
Don't Use Tiny Type
It's very hard to read text on computer screens. Make everything bigger than you would in print. And make sure your type contrasts well with the background. If you want somebody to notice something, make it easy to read.
Viewers won't intuit your site.
You may design an intricate structural hierarchy, boasting a home page and various theme pages, subtheme pages, and so on. But once people have left the first page and begin frantically clicking around, they've lost all concept of your organization; all of the pages are completely equal. Your linear hierarchy is reduced to a horizontal mess in a matter of seconds.
So it's very important to always tell the viewer exactly where they are and how to get back. Boil your site down to as few pages as possible. If a user gets lost on your site, he'll click out and never come back. Run the risk of overexplaining. Make it painfully simple.
Clearly structured sites hold your hand all the way through. They're simple, with consistent, clear navigation on every page. This seems obvious, but look at how many sites change the look and feel of the navigation throughout. The user is never sure that he hasn't been flungto some other site. Bad idea.
One secret to navigation is to limit the use of hypertext. The overuse of hot links will result in at least mild befuddlement. It's used for the sake of novelty way too often.
Perhaps the most neglected aspect of navigation is allowing people to get back to where they were. Too often the viewer is reduced to wildly punching the browser "back" button. The "go back" button is the hallmark of a badly designed site.
In the 1950s, robots were to become mechanical humans.
Forty years later, we realize we were wrong. We've learned they are only good for individual tasks, and a lot of them together can sort of imitate what humans do.
The robotics part of personalization is agenting software such as Firefly (www.agentsinc.com). This application builds a profile of customers or users preferences on a particular subject. For example, music. You rank albums, groups, and songs according to what you like. After you rank several lists, the software builds a profile of your taste and compares it to others. Then, like the old-fashioned clerk in a record store, it recommends what you might like‹that is, what is missing from your list that appears on the lists of those with similar profiles.
But can these programs be scaled? A thousand profiles work well, but when you increase to a million profiles, they start to gray out until they are uselessly general.
Pointcast (www.pointcast.com) allows users to select categories of news, enter the stocks they want to track, while the screen dynamically conforms to these preferences. While this may be intriguing, it only solves part of the problem. If you're normally not interested in weather, for instance, you may not be alerted that there is a tornado heading your way. There needs to be a section added to these pages called Randomly Interesting Stuff.
This, of course, will need to be done by actual humans. But it is important. In fact, people tend to be most interested in things they didn't expect to learn about. In the old days, we called that news.
If you think about it, there are basically three activities on the Net:
Media, which is what most Web sites have to do with;
Communications, such as chats and e-mail;
Transactions, which are really the primary form of business on the Internet.
The Internet is not about brand advertising or promotion that's done better in television and print. Nor is the Internet about shopping‹it's really about buying! Direct marketing can work beautifully on the Internet if we can figure out how to hook up a customer directly to a product and make a sale. If you can hook up a personal profile to a product, you create the most welcome kind of marketing. People are always more interested in getting direct mail if it's information they're interested in. If you're in the market for a car, the glossy Mercedes brochure looks a lot better.
First Virtual (www.fv.com) accepts credit card information over the phone rather than the Internet, and then issues customers a virtual PIN, which can be used at any First Virtual seller's Web site. The Java-based site connects the customer directly to the transaction, without having to wade through forests of screens. And perhaps most impressive in the long run is the fact that the vendor's server with the credit card information is not connected to the Web.
While some react with fear and loathing to the idea of an online bank, others see it as an empowering tool‹and guarantee healthy competition. The fundamental thing about the Web‹it's shocking how often Web developers forget this is that the customer is in charge.
It's a culture clash.
Web sites often house varied talent under the same roof: software people, cable people, marketers, salespaople, editors, directors, writers, designers, etc., etc. So how do we work together?
Leave your assumptions at the door. Our impressive track record in software, design, or editing doesn't mean beans here. It's time to throw away your Oscars, Emmys, and other meaningless awards.
Build a team with representatives from each of the converging industries. And, if possible, hire people who have experience in more than one field. It is equally important to pull from different age groups and cultural backgrounds. These sorts of mixtures provide rich experience and solid groundwork for cross-training.
Start cross-training. Get the software engineers to try to produce content, get the editors to try to write code. Get everybody to think about marketing. Team members should spend time in other areas, and begin to comprehend the scope of the project.
Set up a physical environment that allows a completely open process. Don't have the content department walled off from the marketers. Nobody ever made a blockbuster working out of a cubicle. In fact, the old-fashioned newsroom had several virtues: the lack of claustrophobic work stations, and the open space where you didn't have to call someone on the phone to see if they were there.
Fund the content. In Hollywood, the preprint costs of making a picture are about half the total expenditure. But in most Internet operations, it's a woeful 25 percent or less! After we're all over the flash of this new medium, content is what we have left. Spending up front may be painful, but it's the only way it will work. It takes money to build a new medium.
Keep your team from slinking back to what they're comfortable with. We have to leave our comfortable homes what we're used to and stride bravely into this new frontier. It won't be easy; we haven't even built this place yet. What we're doing is colonizing Mars and we don't even have the damn oxygen in!
A few shots at the tin cans of tomorrow.
Spring 1997 Personalization
The inevitable move toward totally custom sites for the viewer takes over the Net.
Fall 1997 Transactions
The primary form of business on the Net explodes as we realize how to hook a customer directly to a product.
Spring 1998 Broad-band
There are a million subscribers to cable modem services, which allow rich media like video to flow right into everyone's houses.
Fall 1998 Entertainment
The entertainment mavens notice that video on the screen seizes everybody's attention. A rush of big players threaten to decimate the independent Web we know and love.
Spring 1999 Databases
Finally, database technology catches up to the Net. The quality of the database becomes all-important; it becomes understood that this is the only limit on transactions, personalization, and video serving.
Fall 1999
A New Medium
We see the first real results of media integration. Finally, the Net moves past being "radio with pictures" and establishes itself as something totally different: a new medium!
Spring 2000 Long Live Print!
Amid the general millennium fever there arises a crisis of confidence. The numerous companies that poured their every last resource into online notice that print did not die. In fact, print is back! There will be great weeping and gnashing of teeth.
Fall 2000 Ubiquity
The Net is now in every room. The successor to Java is running every electrical appliance that you own. Bill Gates is nowhere to be seen.
2001 Content
The entire industry realizes that the Internet means nothing without good stories, good personalities, and good direction. Content is king. Soon thereafter, a giant monolith is found on the other side of the moon, and, after the expedition to Mars, none of this matters.
Roger Black is one of the most influential designers in the magazine and newspaper world. As art director of the New York Times and Rolling Stone, Black developed an immediately identifiable style: sparse, clean layouts, a strong use of classic fonts, a style that made editorial its priority. During this period, Newsweek, Esquire, Foreign Affairs and Ad Age, were redesigned by Black.
As a pioneer in desktop publishing, Black was a natural candidate for the Internet. In 1994, Black cofounded (with David Berlow and Jock Spivy) the Interactive Bureau. With successes as the Discovery Channel Online and USA Today, IAB is making its presence felt on the Web.